Spatio-temporal Distribution and Transport Pathways Analysis of Sand and Dust Weather in North China

Abstract

This study commences by extracting hourly data on PM10 and PM2.5 concentrations to discern occurrences of sand and dust events in the North China region spanning from 2015 to 2023. Subsequently, the HYSPLIT model is employed to precisely locate the sources and track the migration routes of specific representative sand and dust weather occurrences. Additionally, the study investigates the interplay between meteorological factors and dust events to elucidate the triggering mechanisms of these phenomena. Building on these findings, used to forecast PM10 concentrations for North China through random forest model. The findings indicate In March 19 to 23, 2023, North China witnessed dust events, with the episode from being the most intense and widespread in recent years. An analysis of the paths and sources of this sand and dust event revealed that the severe sandstorm in North China was the result of a synergistic effect of dust sources from Mongolia and Northwest China. The study identified a combination of meteorological factors—maximum wind speed > 6.1 m/s, relative humidity < 56%, and solar radiation > 128 w/m2—as having the most significant impact on sandstorm in North China. Leveraging these research outcomes, the study established an hourly PM10 prediction model for North China using random forest approach, with the highest accuracy for 1–3 hour forecasts. This study has advanced our understanding of dust and sand weather events to a certain extent,, furnishing a theoretical and scientific basis for the effective management of sand and dust in ecologically vulnerable regions.